PROJECT SUMMARY Neonatal infection, and in its severest form, sepsis, are leading causes of morbidity and mortality in the neonatal population, accounting for 24% of newborn deaths worldwide. Despite advances in neonatal care, timely identification of an infected newborn remains a significant diagnostic challenge. Early clinical signs of sepsis are subtle or more often mirror symptoms commonly seen in the premature newborn (i.e. apnea). Decades of research on inflammatory biomarkers has determined that an ideal infection screening platform must be designed to serially and simultaneously monitor multiple biomarkers. However serial serum sampling in the newborn is impractical, noxious and invasive. A safe alternative to repeated blood draws would be to quantify biomarker levels through noninvasively obtained saliva samples. The overall goal of this research application is to pair the expertise of MPIs Drs. Jill Maron and David Walt to translate the first noninvasive test to simultaneously quantify six inflammatory biomarkers in neonatal saliva from serial time points to improve infection-screening accuracy and reduce unwarranted antibiotic exposure in the newborn. The Maron Laboratory at Tufts Medical Center (TMC) has spent the last decade advancing the field of neonatal salivary diagnostics, including being the first to demonstrate the clinical utility of neonatal salivary c-reactive protein (CRP) quantification. In parallel, the Walt Laboratory at Harvard Medical School has invented multiplexed Single Molecule Array (SiMoA) technology capable of quantifying multiple proteins in saliva from a single sample source at a femtoscale level. Together, we have optimized and adapted the SiMoA platform to successfully quantify six inflammatory biomarkers in neonatal saliva (CRP, procalcitonin, tumor necrosis factor- alpha [TNF-?], and interleukins [IL] 1?, 6, and 8). In our proposed prospective, observational trial, with training and validation cohorts, we have paired with international experts in neonatal infection and immune response, Co-Investigators, Dr. James Wynn (University of Florida, Gainesville [UF]) and Dr. Joseph Bliss (Women and Infants? Hospital [W&I]) to develop and validate a predictive model of neonatal sepsis. In Aim 1, pertinent clinical and demographic data will be combined with salivary biomarker signatures of 2,250 infants undergoing a ?rule out sepsis? evaluation at either the TMC or W&I NICUs to develop a predictive model of neonatal infection. In Aim 2, the predictive model will be validated on an independent cohort of 1,750 infants undergoing a rule-out-sepsis at UF. Finally in Aim 3, saliva samples from all uninfected newborns enrolled in Aims 1 and 2 will be used to generate normative salivary values of each biomarker across the neonatal age (24 to 42 weeks) and weight (500 to 4500 g) spectrum, while assessing the potential of these biomarkers to predict other neonatal morbidities associated with inflammation. We aim to enhance the accuracy of sepsis screening, reduce unwarranted antibiotic therapy, and significantly improve neonatal care and outcomes.